The road to the car with a true environmental profile

TU Darmstadt develops intelligent operating concepts for the ‘Vehicle 5.0’


The research association ‘Vehicle 5.0’ focuses on vehicles that continuously learn from operating data and optimize themselves. The knowledge-based approach should also improve the life cycle assessment.

Discussion at the cutaway model of a gear unit: Professor Stephan Rinderknecht, Andreas Viehmann and Arved Eßer (from left to right). Image: Katrin Binner

‘We combine the revolutionary approaches of big data and machine learning with the classical evolutionary methods of vehicle development’: Professor Stephan Rinderknecht, coordinator of the ‘Vehicle 5.0’ research association at TU Darmstadt, formulates a challenging, visionary framework. He believes that new technologies with which gigantic data quantities can be collected, aggregated, and analysed, together with the growing interconnection of vehicles, open new potentials for a paradigm shift in the automotive industry. Digitization offers the opportunity to adapt vehicle properties to the constantly changing driving and environmental conditions during operation, instead of updating them based on fixed reference models and driving profiles as in the past.

The interdisciplinary association anchored at the TU Darmstadt seeks to raise this potential and has already developed concrete application ideas and is currently researching the scientific principles essential for practicability. Experts are not only looking at the technical challenges, but also significant aspects of society as a whole – such as the protection of data and privacy, acceptance issues or consequential costs for the economy and the environment.

‘We want to evaluate the ecological footprint of a vehicle under real conditions,’ summarizes Professor Rinderknecht, head of the Institute for Mechatronic Systems in Mechanical Engineering (IMS), about one of the superordinate objectives. How this can look in practice is shown in the context of the project ‘FahrKLang’. And for that they apply the thus-far unsatisfactory framework conditions: Today, to assess a car’s fuel consumption or CO2 emissions, it undergoes an EU-wide binding measurement process based on standardized driving profiles derived from average data.

This involves two problems: Firstly, these driving profiles rarely correspond with actual usage behaviour. Secondly, an electric vehicle leaves such a test with the label ‘zero CO2 emissions’. Considering this to be misleading, the IMS team is working towards a constantly updated assessment of the environmental impact of a car. The idea is to depict the entire life cycle assessment, from the manufacture of drive components to the generation of electricity and provision of fuel all the way to disposal.

‘For this, we need key figures that are always up to date, and agile driving cycles that continuously adapt based on this real-time data,’ explains Arved Eßer. The IMS scientist shows that a very differentiated picture emerges, particularly when it comes to electric drives. Using statistical data from the Institute for Transport Research at DLR, Eßer and the research team investigated a wide range of drive concepts, optimized the respective drivetrain designs in terms of environmental impact and compared their potentials.

The focus was on the long-distance suitability of the vehicles. He arrived at a differentiated result: ‘Although it is classified as a zero-emission vehicle, today, the pure electric car is counterproductive for actual high operating ranges of 350 kilometres in Germany in.’ Because to make a Battery Electric Vehicle (BEV) suitable for long-distance travel, it requires a very large battery, which on today’s market is not only very expensive but also gives poor life cycle assessment results. At the same time, however, such a pure electric drive architecture offers advantages for ranges of up to 100 kilometres and in city traffic, because there, low battery capacity is sufficient, and a BEV can drive locally without emissions.

‘For long distances, therefore, it makes more sense to additionally install an internal combustion engine and a multi-speed transmission than to invest in very large battery capacities, which are statistically only rarely really used,’ summarised Eßer. For example, the many parallel plug-in hybrids offer the greatest potential, that is to say vehicles that can also be charged at the socket and whose electric motor is supported by a combustion engine if required.

‘We want to drive electrically, but we do not want to throw the attractive technology of the combustion engine overboard for the long haul,’ stresses Rinderknecht. How the two can be combined in as efficient and environmentally-friendly way as possible, the experts have explored in the context of the project ‘DE-REX’. The acronym stands for ‘Two-Drive-Transmission with Range-Extender’. Range extenders are usually combustion engines that drive a generator, which in turn can supply power to the battery and the electric motor. The combustion engine is only used when the battery charge level is too low. The new architecture developed at TU Darmstadt is based on two relatively small electric motors and a combustion engine which, in contrast to conventional range extender concepts, all three can be connected via a simplified automated transmission with the drive shafts of the car.

DE-REX offers many operating modes, which are automatically selected by the operating strategy. For example, at partial load, one electric motor is usually enough; when it comes to full power requirements, both electric motors can be used. When the combustion engine is switched on, the control units choose. While the subtransmission switches the gear of one electric motor, the second drives the car, so that no interruption of traction force occurs. ‘This increases the overall efficiency and drive comfort at the same time,’ explains Andreas Viehmann, responsible for the development of the ‘DE-REX’ at the IMS. In comparison, the prototype, which will remain available for research purposes until 2020, shows promising values: The Darmstadt model was able to reduce the electric power by more than 40 percent versus a comparable serial hybrid drive concept, while increasing the vehicle’s electric range by ten percent.

Both projects show that we are on the right track with our real-driving and knowledge-based methods,’ says Stephan Rinderknecht. For targeted optimization and further development of vehicles, the agile model in which all relevant parameters including the actual ecological assessment are to be incorporated should now be further refined. On the research agenda is also the development of a drive concept for a smart eco-effective universal vehicle that drives locally emission-free in the city, is suitable for long-distances, does not require a new infrastructure and could be integrated in the context of sector-coupling in a smart grid or smart home architecture.

‘Vehicle 5.0’ interdisciplinary

The research association ‘Vehicle 5.0: The knowledge-based automobile of the future’, with its software-based lightweight construction, actual travel optimised drive and accepted autonomous function, comprises three profiling areas of application. Cross-cutting issues are Big Data, Human Factors, Economy and Ecology, and Technical Methods. Vehicle 5.0 projects are currently underway at the Institute of Mechatronic Systems in Mechanical Engineering (IMS), the Institute of Internal Combustion Engines and Powertrain Systems (VKM), the Institute of Automotive Engineering (FZD), the Institute of Electrical Energy Conversion (EW) and the Department of Material Flow Management and Resource Economics (SuR) (all TU Darmstadt). Amongst other organisations, the association cooperates with the National Research Center for Applied Cyber Security (CRISP) and the spin-off Compredict. The projects ‘FahrKLang’ and ‘DE-REX’ were publicly funded by the Federal Ministry for Economic Affairs and Energy.

Read more research stories in hoch³ FORSCHEN 1/2019